A prolonged QRS duration (QRSd) is promising for a response to cardiac resynchronization therapy (CRT). The variation in human body sizes may affect the QRSd. We hypothesized that conduction disturbances may exist in Japanese even with a narrow (< 130 ms)-QRS complex; such patients could be CRT candidates. We investigated the relationships between QRSd and sex and body size in Japanese. We retrospectively analyzed the values of 338 patients without heart failure (HF) (controls) and 199 CRT patients: 12-lead electrocardiographically determined QRSd, left ventricular diastolic and systolic diameters (LVDd and LVDs), body surface area (BSA), body mass index (BMI), and LVEF. We investigated the relationships between the QRSd and BSA, BMI, and LVD. The men's and women's BSA values were 1.74 m 2 and 1.48 m 2 in the controls (p < 0.0001), and 1.70 m 2 and 1.41 m 2 in the CRT patients (p < 0.0001). The men's and women's QRSd values were 96.1 ms and 87.4 ms in the controls (p < 0.0001), and 147.8 ms and 143.9 ms in the CRT group (p = 0.4633). In the controls, all body size and LVD variables were positively associated with QRSd. The CRT response rate did not differ significantly among narrow-, mid-, and wide-QRS groups (83.6%, 91.3%, 92.4%). An analysis of the ROC curve provided a QRS cutoff value of 114 ms for CRT responder. The QRSd appears to depend somewhat on body size in patients without HF. The CRT response rate was better than reported values even in patients with a narrow QRSd (< 130 ms). When patients are considered for CRT, a QRSd > 130 ms may not be necessary, and the current JCS guidelines appear to be appropriate.